Dataset Information

A probabilistic pathway score (PROPS) for classification with applications to inflammatory bowel disease.

ABSTRACT:

Summary

Gene-based supervised machine learning classification models have been widely used to differentiate disease states, predict disease progression and determine effective treatment options. However, many of these classifiers are sensitive to noise and frequently do not replicate in external validation sets. For complex, heterogeneous diseases, these classifiers are further limited by being unable to capture varying combinations of genes that lead to the same phenotype. Pathway-based classification can overcome these challenges by using robust, aggregate features to represent biological mechanisms. In this work, we developed a novel pathway-based approach, PRObabilistic Pathway Score, which uses genes to calculate individualized pathway scores for classification. Unlike previous individualized pathway-based classification methods that use gene sets, we incorporate gene interactions using probabilistic graphical models to more accurately represent the underlying biology and achieve better performance. We apply our method to differentiate two similar complex diseases, ulcerative colitis (UC) and Crohn's disease (CD), which are the two main types of inflammatory bowel disease (IBD). Using five IBD datasets, we compare our method against four gene-based and four alternative pathway-based classifiers in distinguishing CD from UC. We demonstrate superior classification performance and provide biological insight into the top pathways separating CD from UC.

Availability and implementation

PROPS is available as a R package, which can be downloaded at http://simtk.org/home/props or on Bioconductor.

Contact

rbaltman@stanford.edu.

Supplementary information

Supplementary data are available at Bioinformatics online.

SUBMITTER: Han L

PROVIDER: S-EPMC5860179 | biostudies-literature | 2018 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A probabilistic pathway score (PROPS) for classification with applications to inflammatory bowel disease.

Han Lichy L Maciejewski Mateusz M Brockel Christoph C Gordon William W Snapper Scott B SB Korzenik Joshua R JR Afzelius Lovisa L Altman Russ B RB

Bioinformatics (Oxford, England) 20180301 6

<h4>Summary</h4>Gene-based supervised machine learning classification models have been widely used to differentiate disease states, predict disease progression and determine effective treatment options. However, many of these classifiers are sensitive to noise and frequently do not replicate in external validation sets. For complex, heterogeneous diseases, these classifiers are further limited by being unable to capture varying combinations of genes that lead to the same phenotype. Pathway-based ...[more]

PMID: 29048458

Similar Datasets

Project description:BackgroundCurrent methods for gene-set or pathway analysis are usually designed to test the enrichment of a single gene-set. Once the analysis is carried out for each of the sets under study, a list of significant sets can be obtained. However, if one wishes to further prioritize the importance or strength of association of these sets, no such quantitative measure is available. Using the magnitude of p-value to rank the pathways may not be appropriate because p-value is not a measure for strength of significance. In addition, when testing each pathway, these analyses are often implicitly affected by the number of differentially expressed genes included in the set and/or affected by the dependence among genes.ResultsHere we propose a two-stage procedure to prioritize the pathways/gene-sets. In the first stage we develop a pathway-level measure with three properties. First, it contains all genes (differentially expressed or not) in the same set, and summarizes the collective effect of all genes per sample. Second, this pathway score accounts for the correlation between genes by synchronizing their correlation directions. Third, the score includes a rank transformation to enhance the variation among samples as well as to avoid the influence of extreme heterogeneity among genes. In the second stage, all scores are included simultaneously in a Bayesian logistic regression model which can evaluate the strength of association for each set and rank the sets based on posterior probabilities. Simulations from Gaussian distributions and human microarray data, and a breast cancer study with RNA-Seq are considered for demonstration and comparison with other existing methods.ConclusionsThe proposed summary pathway score provides for each sample an overall evaluation of gene expression in a gene-set. It demonstrates the advantages of including all genes in the set and the synchronization of correlation direction. The simultaneous utilization of all pathway-level scores in a Bayesian model not only offers a probabilistic evaluation and ranking of the pathway association but also presents good accuracy in identifying the top-ranking pathways. The resulting recommendation list of ranked pathways can be a reference for potential target therapy or for future allocation of research resources.

Dataset Information

A probabilistic pathway score (PROPS) for classification with applications to inflammatory bowel disease.

Summary

Availability and implementation

Contact

Supplementary information

Publications

A probabilistic pathway score (PROPS) for classification with applications to inflammatory bowel disease.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets